clock menu more-arrow no yes

Filed under:

Strength of Schedule

New, 18 comments

With little else to do on yet another day without Brewers baseball, I took a look at the rest of the regular-season schedule. Here's what we've got:

Team       Games   Home    Away    WP      3WP
COL     6       3       3       0.500   0.504
ARI     7       4       3       0.522   0.457
SF      6       3       3       0.442   0.502
CIN     13      6       7       0.409   0.475
STL     10      6       4       0.471   0.460
NYM     3       3       0       0.552   0.546
PHI     3       3       0       0.500   0.485
HOU     9       3       6       0.438   0.443
CHC     3       0       3       0.506   0.529
PIT     6       3       3       0.455   0.424
ATL     4       0       4       0.528   0.516
SD      4       4       0       0.563   0.540
By way of clarification, "WP" is the team's actual winning percentage to date, while "3WP" is the team's third-order winning percentage, as included in BP's adjusted standings report. If you weight those winning percentages by the number of games we play against each team, the story is a happy one: by regular winning percentage, the average opponent is .461; by BP's method, average is .466. I'll take either one.

I did the same exercise for Cubs, and found that while they have a pretty easy schedule too (it's very similar to ours), it's a little tougher. The main difference is that they have a four-game set against the Dodgers, while we do not. Their opponents are, on average, .477 (regular) and .482 (BP).

Then I went a little crazy. Honestly, when I sat down to do this, I was going to stop right there. (Actually, I wasn't going to do the Cubs at all.) But I ended up writing a script to simulate the rest of the season for the Brewers and Cubs, much like the BP postseason odds report. Only for just one or two teams.

Each time I did this, I did it twice: once for actual winning percentages, and once for third-order WPs. For the Brewers, using actual WP, the median number of wins is 92. I ran a million simulations, and we come out below 85 wins only 3.5% of the time, while we get 100 or more 3.9% of the time. The range from 88 to 96 wins is by far the most likely, accounting for about 71% of the possible outcomes.

Using third-order winning percentages, the picture is less rosy, but still okay. The median moves down to 90; there's less than a 10% chance of ending up below 85 wins, and there's about a 4% chance of getting 98 or more wins. About 75% of the outcomes are concentrated between 86 and 95 wins.

I don't want to restart our debates about pythagorean (and variants thereof) vs. actual winning percentage--we've been over that territory before--but for unrelated reasons, I think the 'actual' sims are probably more accurate. Not because you should use actual numbers (I don't think that) but because none of the numbers take into account the true talent level of the second-half version of the Brewers.

For instance, the first-half Brewers didn't have Corey Hart for a couple of weeks. We had no Ryan Braun for several weeks, and no Yovani Gallardo until recently. If Jeff Suppan's history is any indication, he'll pitch better from here on out than he has to this point. And it seems likely that Doug Melvin will make some kind of move that, presumably, will improve the team. There are counterarguments, of course: we're going to be missing Bill Hall for a while, and some players (notably Braun) probably won't keep up their current pace, but I think the reasons to think the Brewers will be a better second-half team are stronger than those to disagree.

Finally, I did all of those calculations with the Cubs, too, and then ran 1,000,000 simulations of the rest of the season to see whether the Cubs or Brewers would win the division. (I assumed away the possibility that any other team would win the division.)

Using either actual WP or third-order WP, the numbers are almost identical. The Brewers win 64.5% of the time, the Cubs win 29.5% of the time, and the teams tie about 6% of the time. I'm a bit baffled as to why those numbers differ substantially from the BP odds report (they give the Brewers a 71% chance of winning the division); while I'm not using Clay's exact method, I'm getting pretty close.

I wish the Brewers had done a few things differently to this point so that the division were a bit more locked up--heaven knows it could be. But with plenty of reasons to be optimistic for the second half, it's certainly Milwaukee's division to lose.